Currently, 3rd generation cellular communication systems are being rolled out to further enhance the communication services provided to mobile phone users. The most widely adopted 3rd generation communication systems are based on Code Division Multiple Access (CDMA) and Frequency Division Duplex (FDD) or Time Division Duplex (TDD) technology. In CDMA systems, user separation is obtained by allocating different spreading and/or scrambling codes to different users on the same carrier frequency and in the same time intervals. This is in contrast to time division multiple access (TDMA) systems, where user separation is achieved by assigning different time slots to different users.
In addition, TDD provides for the same carrier frequency to be used for both uplink transmissions, i.e. transmissions from the mobile wireless communication unit (often referred to as wireless subscriber communication unit) to the communication infrastructure via a wireless serving base station and downlink transmissions, i.e. transmissions from the communication infrastructure to the mobile wireless communication unit via a serving base station. In TDD, the carrier frequency is subdivided in the time domain into a series of timeslots. The single carrier frequency is assigned to uplink transmissions during some timeslots and to downlink transmissions during other timeslots. An example of a communication system using this principle is the Universal Mobile Telecommunication System (UMTS). Further description of CDMA, and specifically of the Wideband CDMA (WCDMA) mode of UMTS, can be found in ‘WCDMA for UMTS’, Harri Holma (editor), Antti Toskala (Editor), Wiley & Sons, 2001, ISBN 0471486876.
Referring now to FIG. 1, a message sequence chart 100 of an Intra-LTE handover procedure in asynchronous network is illustrated, as agreed in R2-072847, ‘Draft1 minutes’ of the 58th TSG-RAN WG2 meeting, Kobe, Japan, 7-11 May 2007. The approach agreed at RAN2#58 describes communications between a user equipment (UE) 110, a source eNodeB 115 and a target eNodeB 120. The approach facilitates UE access to a target cell using a contention-free procedure with dedicated resources. According to the agreed procedure, the UE performs signal quality measurements and transmits these measurements, in message 130, to the source eNodeB 115. The source eNodeB 115 then initiates a handover (HO) process 135 and transmits a HO request message 140 to the target eNodeB 120 (over the UE radio access network (RAN)).
The target eNodeB 120 performs admission control 145 and allocates a dedicated preamble for RACH access in the target cell during the admission control process 145. The allocated dedicated preamble message is sent from the target eNodeB 120 to the source eNodeB 115 in a handover (HO) request acknowledge (ACK) message 150 using a new Cell specific Radio Network Temporary Identifier (C-RNTI) and thereafter to the UE 110 in a handover (HO) command message 155.
The UE 110 then transmits an acknowledge message 160 to the source eNodeB 115, which prepares the data for forwarding in both an uplink (UL) and downlink (DL) direction 165. The prepared data is then forwarded 170 from the source eNodeB 115 to the target eNodeB 120. The UE 100 is then able to perform a RACH access with the target eNodeB 120 using the allocated dedicated preamble 175, who replies with a RACH response indicating a timing advance (TA) and UL grant details 180. The UE then sends a HO confirmation message 185 to the target eNodeB 120, which replies with a further ACK message 190. Thus, this is an asynchronous handover, as the UE is not synchronized to the new cell prior to the access.
Preamble space is partitioned into two parts, namely as dedicated preambles and random preambles. For normal RACH access, UE 110 randomly selects a preamble from the random preamble portions and transmits the selected preamble over a non-synchronous RACH channel 175. Only the preambles within the random preamble portion needs to be broadcast in the cell. Dedicated preambles are always allocated to the UE 110 by the network (eNodeB). Referring now to FIG. 2, a message sequence chart 200 illustrates a case where there is no available preamble to be dedicated to the UE 210 access, or the target eNodeB 220 does not allocate a dedicated preamble for RACH access in the target cell. Here, the UE 210 accesses the target eNodeB 220 via contention-based non-synchronous RACH access, and in response thereto the target eNodeB 220 transmits a HO request ACK message 250 to the source eNodeB 215, which in turn transmits a HO command 255 to the UE 210, without carrying a dedicated preamble.
With respect to the contention aspect used in this procedure, for instance when two or more UEs have selected the same preamble, the UEs will be listening to the same RACH response 180. However, the target eNodeB 220 is only able to detect the strongest signal. Hence, the TA will be calculated for the UE with the strongest signal. In this case, all the contending UEs receive the TA and assume that it is their own TA. After performing timing alignment, the contending UEs then transmit their unique identifier on the scheduled resources. Note that, typically, more than one UE is transmitting on the same UL resources, thereby causing further collision.
However, if the eNodeB receives and decodes the signal 175 transmitted by a UE correctly, the eNodeB sends the UE's unique ID in message 180. All the contending UEs listen to the transmitted message in message 180. If the unique ID matches the UE's respective identifier (ID), that UE has successfully accessed the cell. In this case, other failed UEs re-start the RACH access by repeating the procedure from message 275 by transmitting another randomly selected preamble. If the scheduled transmission message 280 is unable to be received correctly by the target eNodeB 220, say due to the resource collision, the contention message of step 285 is not possible. In this case, the UEs re-start the RACH procedure after expiry of a timer.
Also, if the dedicated preamble (non-contention) based access fails, the UE 210 will access the target cell using a randomly selected preamble on a non-synchronous RACH channel 275. In this case, a UE 210 transmits a scheduled transmission 280 to the target eNodeB 220, who responds with contention resolution step 285, prior to the final confirmation and ACK transmissions 185, 190. Thus, the agreed asynchronous HO procedure requires the UE 110, 210 to access the new cell on a non-synchronous RACH channel. The HO load contributes significantly to the total RACH load. According to the RACH load analysis provided by Samsung™ in R2-07025, in the document titled ‘LTE cell load/RACH load estimations’, Samsung, RAN2#56bis, Sorrento, Italy, 15-19 Jan. 2007, 50-70% of the RACH load is caused by cell access after handover.
It is known that a reduction of RACH load in such handover procedures is always beneficial from a radio efficiency perspective.
The aforementioned known prior art deals with handover in asynchronous networks, where the DL timing is not synchronized. In contrast, in a synchronized network, the downlink (DL) transmissions are synchronized (i.e. DL frame timing occurs at the same time at different eNodeBs). However, the timing advance (TA) is a timing alignment that is needed by the UE to adjust for UL transmissions.
In synchronized handover, a UE is capable of obtaining UL synchronization to the new cell prior to the cell access in the new cell. In a synchronous network the UE is able to calculate the timing advanced based on the TA in the source cell and the time difference between the signals received from the source and the target cells. In an asynchronous network, if the time difference between the source and the target cell is given to the UE, the UE is capable of calculating the TA.
To summarize, based on a particular location of a UE and a speed of the UE, the waveforms transmitted by the UEs (UL transmission), if transmitted at the same time, will be received by the eNodeB at different times. To correct this (i.e. to ensure that the UL transmission will be received by the eNodeB in the same time slot), the eNodeB orders each UE to transmit at different times, which is referred to as Timing Advance. Thus, UL time alignment is required by the UE in both synchronized and asynchronized networks.
In normal operation, in wireless networks, the TA is calculated by the eNodeB based on the transmitted signal by the UE. The calculated TA is then sent to the UE, so that the UE is able to accordingly adjust its UL transmission timing. In synchronized networks, the TA in the target network (during a handover (HO) operation) may be calculated by the UE without any involvement of the target eNodeB. This is achieved based on a received timing difference between the signal from the source eNodeB and the target eNodeB and the current TA (known to the UE) employed in the source cell.
In an asynchronized network, this calculation is only possible if the UE is provided with the DL frame time differences between the source cell and target cell.
Thus, to clarify handover procedure in a synchronised network, the UE is able to calculate the timing advance to the new target eNodeB based on a relative time difference between the received signals from the new and old cells. This is the mechanism used in known TDD-UMTS networks. Therefore in a synchronised network it is possible to obtain synchronisation with the new cell, prior to access, and thus to avoid access in the new cell via a non-synchronous channel. A handover where the UE obtains the UL synchronisation information for the new cell, prior to cell access, is termed a ‘synchronous handover’.
Referring now to FIG. 3, a message sequence chart 300 illustrates a case of a synchronous handover procedure. In an MBMS Single Frequency Network (MBSFN) operation (regardless of whether the network is time division duplex (TDD) or frequency division duplex (FDD)), the cells are DL frame synchronized.
In a synchronous network, timing advanced in the target cell can be calculated by the UE 310 simply, based on TA for the source eNodeB 315 and relative time difference between received signals from the target eNodeB 320 and source eNodeB 315. An algorithm similar to that used in the TDD-UMTS system can also be designed for the 3GPP LTE Standard. The UE 310 is able to gain UL time synchronization to the target cell prior to access, and avoid non-synchronous RACH access, and hence reduce the RACH load in the target cell.
In an asynchronous network, timing advanced in the target cell can be calculated by the UE 310 only if the DL frame differences between the source cell and the target cell are known by the UE.
One possible way of avoiding RACH access in the target cell is for the target eNodeB 320 to allocate UL signaling channel (SCH) resources with the allocation signaled via the source eNodeB 315 over the HO request ACK message 350 and the subsequent HO command 355.
The signaling flow involved in such a UE based TA calculation scheme in a wireless network is illustrated in the message sequence chart 300 of FIG. 3. Thus, after receiving the HO request 140 from the source eNodeB 315, the target eNodeB 320 allocates a C-RNTI and assigns resources on an HO request ACK UL-SCH message 350 to the source eNodeB 315. After receiving the Handover command 355, the UE 310 detaches from the existing cell and synchronizes to the new (target) cell. Then a Handover confirm message 185 is sent over the allocated UL-SCH resources.
This procedure is simple. However, the allocation of UL-SCH resources in this manner may result in a waste of radio resources. Thus, in this case, the HO command may take several hybrid automatic repeat request (HARQ) transmissions to be correctly received by the UE 310. In this manner, the UL-SCH resources may need to be reserved for the use of HO confirm message 185 for a longer duration, which is not desirable from radio efficiency perspective.
Alternatively, a start timer for the transmission of a HO confirm message 185 may be set by the target eNodeB 320. In this case, the start time should be set taking into account a worst case delay over an .times.2 interface, which is the interface between two eNodeBs, and maximum HARQ transmission delay. Thus, the HO interruption time 395 may be un-necessarily large for some transmissions.
A yet further method for synchronous HO procedure in wireless networks, which has been proposed by Motorola™ in R2-070214 in a document titled ‘Contention and contention-free intra-LTE handovers’ is illustrated in the message sequence chart 400 shown in FIG. 4.
Here, after receiving the HO request 140 from the source eNodeB 415, the target eNodeB 420 allocates a C-RNTI and assigns dedicated resources for CQI reporting and scheduling request channels in a HO request ACK message 450. The target eNodeB 420 conveys this information to the UE 410 via the source eNodeB 415 in HO command message 455.
After synchronizing to the target eNodeB 420, the UE 410 accesses the target cell by sending a CQI report 475 or scheduling request (SR) message on the allocated dedicated resources. The target eNodeB 420 thus allocates dedicated resources to the UE 410 in a layer-1/layer-2 control channel (UL grant) message 480.
Similar to the procedure shown in FIG. 3, the message sequence chart 400 of FIG. 4 also relies on UE communication on dedicated resources in the target cell. However, as shown, CQI/SR resources 450, 455 are used in FIG. 4 instead of UL-SCH resources 350, 355 in FIG. 3.
Thus, CQI/SR resource space can be visualized as code and time space, where CQI/SR reporting may also be either periodic or triggered. In a case of periodic reporting, a UE 410 will be allocated code and time resources in a periodic pattern. From a perspective of the UE 410, the allocated CQI/SR channel is dedicated to the UE 410 at a given time, hence allowing contention-free CQI reporting and scheduling request.
CQI reporting assists the link adaptation on DL transmission. Hence, the CQI reporting is only useful for UEs in active transmission. The UE 410 is able to be categorized into two modes depending on their activity level in LTE_Connected states. If the UE 410 is in an RRC_Connected state, and it is actively involved in the communication, the UE 410 is considered to be in a short_DRX or continuous state. For UEs that are RRC_connected, but not actively involved in a communication, the states are considered as long_DRX states.
In practice, there are a huge number (several thousands) of UEs in long_DRX state, whilst only about 200-400 UEs may be in short_DRX or continuous reception mode. Thus, in practice, the proposed approach in FIG. 4 is inefficient in allocating a dedicated CQI/SR channel for UEs operating in long_DRX state. Furthermore, the allocation of CQI channel resources may also show some level of radio resource wastage, due to the large message length used in the HO command.
Furthermore, allocation of dedicated CQI/SR resources in advance may also result in a waste of radio resources, in case the UE 410 takes long time to access the new (target) cell.
Thus, current handover techniques, particularly those suggested for synchronous handover in wireless networks are suboptimal. Hence, an improved mechanism to address the problem of synchronous handover in a wireless networks would be advantageous; particularly one that reduces or removes RACH load.